Multiple drug resistance (MDR) has become a central theme in the study if innate and acquired drug resistance of tumor cells. Although such resistances may be multifactorial in nature, expression of the human mdrl gene has been shown to correlate with MDR in numerous cell lines and has been found at high basal levels in certain normal human tissues. Genes associated with the MDR phenotype have been cloned and characterized from several species. The mdrl gene codes for a high molecular weight membrane associated glycoprotein (PGP) which evidently mediates the extrusion of drugs subject to cross-resistance in the MDR phenotype. In the human, expression of the mdrl gene correlates with MDR in several cell lines. Moreover, the MDR phenotype can be conferred upon a recipient cell line by transfection of the mdrl gene. On the other hand, partial reversal of the MDR phenotype has been accomplished by the administration of calmodulin inhibitors (eg verapamil). These results suggest that circumvention of the MDR phenotype in tumors might lead to better antitumor effects of chemotherapeutic drugs which are extruded from cells by the gene products of mdrl. A more direct approach to ablating expression of the mdrl gene would be to use antisense. A shuttle vector id described here which should enable the expression of relatively high levels of antisense mdr RNA. The vector makes use of a promotor derived from GRP78 gene, which is inducible by calcium ionophores in human cells. MDR-KB cell lines expressing various levels of the mdrl gene will be transfected with this inducible mdr-antisense vector, in combination with an independent selectable marker, Transfected clones expressing differing levels of anti- mdr mRNA will be isolated, and changes in levels of mdr mRNA and PGP gene product as well as changes in drug sensitivity will be determined. Since these cell lines are well characterized, the results should provide information as to the feasibility of overcoming MDR by antisense. In addition parental drug-sensitive KB cell populations will be transfected with the anti-mdr mRNA expressing vector. The rate of mutation to the MDR phenotype (ie., coordinate resistance to colchicine, adriamycin and vinblastine accompanied by PGP overexpression) will then be estimated for these cells in comparison to untransfected controls. These studies should allow estimates of the levels of antisense mRNA needed to overcome relative levels of gene expression, and also evaluation of the feasibility of diminishing the appearance of the MDR phenotype in drug naive cells.